What would you like to do next?
2–6 past start dates in, your spread out.
Cycle length calculator →Now the days-late logic applies.
Late period calculator →The honest fertile-window range for irregular cycles.
Ovulation calculator →If your cycles vary by a week or more, single-date predictions fail you. This calculator gives an honest date range instead — from your shortest and longest recent cycles, or straight from your past period dates.
Predicted period days, fertile window and ovulation — based on your average cycle.
What counts as irregular: a shortest-to-longest spread beyond 7–9 days across a year (FIGO), or cycles persistently outside 24–38 days. How to predict: a window, not a day — earliest = last period + shortest cycle, latest = last period + longest. How common: 14–25% of women of childbearing age (NICHD). You are far from alone.
Irregular is not “my period never comes on the same date” — nobody's does. The FIGO line: if your shortest and longest cycles over a year differ by less than 7–9 days, you count as regular, even if that means 26 days one month and 33 the next. A wider spread, or cycles consistently outside the adult normal frequency of 24–38 days, is irregular.
The 7–9 days is really two numbers — FIGO sets the line by age:
Two life stages get a pass by physiology: the first years after the first period (ACOG considers 21–45-day cycles normal for teens) and perimenopause. New or widening irregularity outside those stages is what deserves attention.
Every standard calculator does the same thing: last period + average cycle = one date. That works when cycles are steady, because the variable half of the cycle — the follicular phase, period start to ovulation — only wanders a little. With irregular cycles it wanders a lot, and averaging hides it:
One more piece of evidence this is a real pain: across period-tracker communities, the most common complaint is apps built on a 28-day average confidently predicting dates that arrive weeks off for irregular users — with no way to correct them. A range never lies to you like that. The honest prediction is the range. One more thing worth understanding: uncertainty compounds — the window for your next period is your spread wide, but two cycles out it can be nearly double. That is a property of irregular cycles, not a flaw in the tool.
| Your shortest–longest | Spread | Reading |
|---|---|---|
| 27–31 days | 4 days | Regular by FIGO's parameter — a single average would serve you fine |
| 26–35 days | 9 days | Borderline — at the edge of the regularity band |
| 24–40 days | 16 days | Clearly irregular — range prediction is the only honest format |
| 21–45+ days | 24+ days | Strongly irregular; check whether a life stage or a cause explains it |
The two ends are not error margins — they are your own shortest and longest recent cycles mapped onto dates. A narrow window (a few days) means your pattern is steady and the estimate is nearly a date. A wide one means your body genuinely doesn't keep a schedule, and no app can conjure precision you don't have — anyone who sells you a single confident date for a 24–40-day pattern is guessing.
The wider your window, the more it pays to stack a body signal on top of the calendar: a clear, stretchy discharge or an LH strip narrows the fertile side of the guess far better than more averaging (the ovulation calculator covers that timing logic). Signals are clues, not proof: a positive LH strip or fertile-type mucus shows the body is gearing up to ovulate, but does not guarantee that ovulation happened.
| Scenario | Inputs | Prediction | Reading |
|---|---|---|---|
| Mildly irregular | Last period = day 0; shortest 26, longest 31 | Window: day 26–31 | 5-day window; expect the period any day inside it, no action needed |
| Clearly irregular | Last period = day 0; shortest 24, longest 40; today = day 33 | Inside the window (24–40) | “Late” does not apply yet — the window has not closed |
| Past the longest cycle | Longest 38; today = day 45 | 7 days beyond the window | Now late-period logic applies: test if pregnancy is possible, escalate at 90 days |
The most common cause of infrequent ovulation. WHO estimates 6–13% of reproductive-age women, with up to 70% undiagnosed. Long-standing irregularity since the teens is a classic clue.
Thyroid hormones feed into the cycle-regulating axis. Watch for fatigue, temperature sensitivity, unexplained weight change.
Each delays or suppresses ovulation, and each delayed ovulation lengthens that cycle. A stressful season often reads as a stretch of irregular cycles.
Teens (first years), postpartum and breastfeeding, and perimenopause all produce physiological irregularity — expected, not pathological.
Natural cycles can take a few months to settle back into a personal pattern.
One weird cycle after a weird month is life, not disease. A repeating pattern is what deserves a closer look.
First day of full flow, not spotting. The period tracker keeps the log on your device, no account.
Three logged gaps give a usable range; six make it reliable. The cycle length calculator computes the spread for you.
A single weird cycle is noise. A repeated pattern is signal.
A pattern: cycle length varies by more than 7–9 days across the year.
A full cycle absent. With irregular cycles, “missed” only starts past your longest recorded cycle.
Missed period calculator →Book an evaluation when:
If you have been brushed off with "it's just hormones" before, go in with the record, not a summary: your logged day-1 dates (the period tracker keeps them on your device), your shortest and longest gaps, how long the pattern has run, and the specific clues you have noticed (skin, hair, weight, energy). A clinician can act on a documented pattern much faster than on "my periods are weird" — and a normal blood panel does not by itself settle a cycle-pattern question.
Many people with irregular cycles still ovulate, just later or less often — but some irregular cycles have no ovulation at all (anovulatory cycles, common with PCOS), so the challenge can be timing, ovulation itself, or both. Your fertile window widens the same way the period window does, and calendar math alone stops being precise: ovulation (LH) test strips detect the surge 24–36 hours before ovulation regardless of cycle length, and are far more useful than any calculator in this situation — though a positive strip shows a surge, not proof that ovulation followed. If cycles are long or unpredictable and you are trying to conceive, raise it with a clinician early rather than waiting out a year of trying. See the ovulation & fertile window calculator for the timing logic.
Already pregnant with irregular cycles? LMP-based dating assumes a 28-day cycle — on your pattern it can be weeks off. Dating is then set by an early first-trimester ultrasound, which is the accurate route; ACOG calls a pregnancy “suboptimally dated” if no ultrasound confirms or revises the due date before 22 weeks 0 days. A rough bridge until then: conception ≈ ovulation day, and the pregnancy week counts from about two weeks before that. Our due date calculator accepts a conception or IVF date directly, and why pregnancy is counted from the last period explains the week-counting rules.
The range method is our methodology, not a medical standard — the assumption above is usually true but not always. It cannot tell you why your cycles are irregular, and it must never be used for contraception. Full formulas and sources: methodology.
Pick the first day of your most recent period — the day full bleeding started, not spotting.
Not sure? Use the built-in helper: enter a few past start dates and we compute your personal average.
Next period date, ovulation day, fertile window and a 6-month forecast — plus a heads-up if you're running late.
When your period actually starts, come back and tap "Log this period". After two or three logs we learn your real cycle length — not the textbook 28 days — and your predictions start using it.
A shortest-to-longest spread beyond 7–9 days over a year (FIGO), or cycles persistently outside 24–38 days. One odd cycle does not make you irregular.
Skip the average. Earliest = last period + shortest recent cycle; latest = last period + longest. That window is the honest prediction — the calculator above does the arithmetic.
Common — an estimated 14–25% of women of childbearing age (NICHD). And normal by physiology for teens in the first years after menarche (ACOG: 21–45 days) and during perimenopause.
Sometimes. PCOS is the most common cause of infrequent ovulation and affects an estimated 6–13% of reproductive-age women (WHO), with up to 70% undiagnosed. Irregularity since the teens plus acne or excess hair growth is worth screening.
Often yes — many irregular cycles still ovulate, just at unpredictable times, and LH test strips help with timing where calendar math gets wide. Some irregular cycles don't ovulate, though, so if cycles are long or unpredictable and you're trying, see a clinician early.
Often, but not always. In many irregular cycles ovulation simply comes late — the half of the cycle before ovulation is what wanders, and the period then follows roughly two weeks after ovulation. But irregular or infrequent periods are also the main sign that some cycles have no ovulation at all (FIGO treats infrequent or irregular bleeding as a sign of an ovulatory disorder; PCOS is an important cause). LH strips and cervical mucus can show the body gearing up, not that an egg was released — a sustained temperature rise afterwards, or a clinician's blood test, is what confirms it. Long or very unpredictable cycles are worth discussing with a clinician.
It depends on what you stopped. Copper IUD: it contains no hormones (it works by releasing copper), so the cycle pattern you see after removal is your own; a copper coil can make periods heavier and more painful while it is in. Hormonal IUD (IUS): it releases progestogen and can make periods lighter, shorter and less painful while it is in, so your usual pattern may look different once it is out. Pill: bleeds on the pill follow the pack, not ovulation, so your first natural cycles after stopping can look different from what you saw on it. With either coil you can get pregnant as soon as it is out (NHS), and after the pill fertility usually returns to previous levels within about a month — so use other contraception straight away if you don't want to conceive, and take a test whenever pregnancy is possible, without waiting for a period. Log day-1 dates from the start; see a clinician if periods haven't come back after 3 missed months (NHS: 3 missed periods in a row), and straight away for very heavy bleeding, severe pain or pregnancy symptoms.
No. With unpredictable ovulation there is no reliably safe window, and calendar estimates must never be used for contraception. If you need contraception, a clinician can help you pick a proven method.
You have no reliable “missed period” date to anchor to, so use your longest recent cycle as the expected date and test from that point — or about 3 weeks after the unprotected sex in question. A negative before that proves little.
Two early cycles is two events, not yet a pattern. Track the start dates; if the shortest-to-longest spread over a year stays under 7–9 days, you still count as regular (FIGO).
Yes. Stress delays ovulation, and since the follicular phase is the variable half of the cycle, each delayed ovulation lengthens that cycle. A stressful few months often reads as a run of irregular cycles.
Often yes — cycles commonly settle after the teen years, postpartum, after stopping hormonal birth control, or when a stressor resolves. Persistent change in the other direction is the one to watch.
Cycles persistently under 24 or over 38 days; a spread beyond 7–9 days across a year; 90 days with no period; new irregularity after stable years; or very heavy bleeding / severe pain.
No. Late is one cycle running long; irregular is a varying pattern. One late period does not make your cycles irregular — see the late period calculator for the single-event scale.